A key problem in organization theory is to suggest new organizational forms. In this paper, I suggest the use of genetic algorithms to search for novel organizational forms by reproducing some of the mechanics of organizational evolution. Issues in using genetic algorithms include identification of the unit of selection, development of a representation and determination of a method for calculating organizational fitness. As an example of the approach, I test a proposition of Thompson's about how interdependent positions should be assigned to groups. Representing an organization as a collection of routines might be more general and still amenable to evolution with a genetic algorithm. I conclude by discussing possible objections to the application of this technique.